HSpice 2004 vs HSpice 2009

Artemij14 · Dec 12, 2009

Hi all

There is a project generator with a YIG-resonator. HSpice 2004 shows the generation of oscillations, but in HSpice 2009 (with the same net-list) simulation results are triangular oscillation (ie there is a mathematical error).

By applying the parameter
.OPTION POST PROBE ACCURATE=1
problem was solved. But no generation of oscillations is observed.

Why are the results so different.

ankitgarg0312 · Dec 13, 2009

hspice has gone dramatic changes in algorithms since 2007 till now...

try below settings
remove .option accurate

add
.option runlvl=5
.option method=trap puretp

if you have any node that need initializations, do a .ic v(node) = v

if above settings gives a triangular wave then add below setting ...

.option delmax = 1/(20*freq)

freq= expected freq of oscillator....

if still does not works, contact solvnet ..
they will help u out ..

Artemij14 · Dec 14, 2009

These options really helped, but generator still does not works.
I found in the output following warning

Warning: 8 subcircuits are ignored because not called by any instance.

Subcircuits not called by any instance:
-------------------------------------------------------------------------------------------
| Location | Subckt |
|------------------------------------------------------------------------+----------------|
(path to my lib)
|------------------------------------------------------------------------------------------|

Could this warning be the cause of incorrect results?

All subcircuits were described as follows
*===================== Subcircuit begin

.SUBCKT GEN (node01 node02 node03 node04 nodeg05 node06)

...
...
.ENDS GEN

*===================== Subcircuit end

Is it correct?

ankitgarg0312 · Dec 14, 2009

Are these subckt getting used in design??

I have doubts on spice format of netlist , I haven't seen parantheses in spice subckt definition.
Please confirm that these paranthesis are valid spice syntax or not.

Added after 3 minutes:

Are these subckt getting used in design??

I have doubts on spice format of netlist , I haven't seen parantheses in spice subckt definition.
Please confirm that these paranthesis are valid spice syntax or not.

Artemij14 · Dec 14, 2009

Even if I don´t use the paranthesis, I get this warning.
As far as i know, this subcircuits definition are valid for HSpice2009. Maybe I am wrong.

ankitgarg0312 · Dec 14, 2009

Best option would be to go to solvnet

Without having a look at both log files I can't comment further

Artemij14 · Dec 15, 2009

Log files:

Code:

***** HSPICE -- C-2009.09 32-BIT (Aug 24 2009) linux ******
 Input File: TG03_Gen_nom.sp
 lic:
 lic: FLEXlm: v8.5b
 lic: USER:   root                 HOSTNAME: localhost.localdomain
 lic: HOSTID: 001d608c9d4c         PID:      30350
 lic: Using FLEXlm license file:
 lic: -1@lcserver
 lic: Checkout 1 hspice
 lic: License/Maintenance for hspice will expire on 28-oct-2010/2009.09
 lic: FLOATING license(s) on SERVER lcserver
 lic:
 init: begin read circuit files,  cpu clock=  0.00E+00
       option search   = /usr/synopsys/C-2009.09/hspice/parts/ad
       option search   = /usr/synopsys/C-2009.09/hspice/parts/behave
       option search   = /usr/synopsys/C-2009.09/hspice/parts/bjt
       option search   = /usr/synopsys/C-2009.09/hspice/parts/burr_brn
       option search   = /usr/synopsys/C-2009.09/hspice/parts/comlinear
       option search   = /usr/synopsys/C-2009.09/hspice/parts/dio
       option search   = /usr/synopsys/C-2009.09/hspice/parts/fet
       option search   = /usr/synopsys/C-2009.09/hspice/parts/lin_tech
       option search   = /usr/synopsys/C-2009.09/hspice/parts/pci
       option search   = /usr/synopsys/C-2009.09/hspice/parts/signet
       option search   = /usr/synopsys/C-2009.09/hspice/parts/ti
       option search   = /usr/synopsys/C-2009.09/hspice/parts/tline
       option search   = /usr/synopsys/C-2009.09/hspice/parts/xilinx
       option icswee   =     1.00
       option parhie   = global
       option runlvl
       option runlvl   =     5.00
       option absi     =    1.000E-09
       option absv     =    1.000E-10
       option post
       option probe
       option method   = trap
       option puretp
       option delmax   =    5.000E-11
 init: end read circuit files,  cpu clock=  2.00E-02 memory=    154 kb
 init: begin check errors,  cpu clock=  2.00E-02
 init: end check errors,  cpu clock=  2.00E-02 memory=    288 kb
 init: begin setup matrix, pivot=     0 cpu clock=  2.00E-02
       establish matrix -- done,  cpu clock=  2.00E-02 memory=    301 kb
       re-order matrix -- done,  cpu clock=  2.00E-02 memory=    306 kb
 init: end setup matrix,  cpu clock=  2.00E-02 memory=    327 kb
 output: TG03_Gen_nom.tr0
 sweep: tran tran0    begin, stop_t=  3.00E-06 #sweeps=**** cpu clock=  2.00E-02
 tran: time= 3.0001E-07 tot_iter=   12127 conv_iter=    6040 cpu clock= 4.90E-01
 tran: time= 6.0001E-07 tot_iter=   24127 conv_iter=   12040 cpu clock= 9.30E-01
 tran: time= 9.0001E-07 tot_iter=   36127 conv_iter=   18040 cpu clock= 1.41E+00
 tran: time= 1.2000E-06 tot_iter=   48127 conv_iter=   24040 cpu clock= 1.85E+00
 tran: time= 1.5000E-06 tot_iter=   60127 conv_iter=   30040 cpu clock= 2.28E+00
 tran: time= 1.8000E-06 tot_iter=   72127 conv_iter=   36040 cpu clock= 2.59E+00
 tran: time= 2.1000E-06 tot_iter=   84127 conv_iter=   42040 cpu clock= 2.84E+00
 tran: time= 2.4000E-06 tot_iter=   96127 conv_iter=   48040 cpu clock= 3.08E+00
 tran: time= 2.7000E-06 tot_iter=  108127 conv_iter=   54040 cpu clock= 3.35E+00
 tran: time= 3.0000E-06 tot_iter=  120127 conv_iter=   60040 cpu clock= 3.59E+00
 sweep: tran tran0    end,  cpu clock=  3.59E+00 memory=    328 kb
>info:         ***** hspice job concluded
 lic: Release hspice token(s)

Code:

************************************************************************
HSPICE RF -- C-2009.09 (Aug 24 2009) 
Copyright (C) 2009 Synopsys,Inc. All rights reserved.
This software and the associated documentation are confidential and
proprietary to Synopsys, Inc. Your use or disclosure of this software
is subject to the terms and conditions of a written license agreement
between you, or your company, and Synopsys, Inc.
************************************************************************

hspicerf 32-bit version C-2009.09 (Aug 24 2009 13:02:33) 

machine info: Linux localhost.localdomain 2.6.18-53.el5xen #1 SMP Wed Oct 10 16:48:44 EDT 2007 x86_64

current time: 16:19:01  12/14/2009

command line: /usr/synopsys/C-2009.09/hspice/linux/hspicerf TG03_Gen_nom.sp

Warning: 8 subcircuits are ignored because not called by any instance.

Subcircuits not called by any instance:
-------------------------------------------------------------------------------------------
| Location                                                              | Subckt          |
|-----------------------------------------------------------------------+-----------------|
| (./Crystall_Analog_V3_3.lib:70)                                       | "TRIG"          |
| (./Crystall_Analog_V3_3.lib:103)                                      | "GEN_M"         |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:101)  | "TK_NL34W25F08" |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:546)  | "TK_NL34W25F32" |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:992)  | "TK_PL30W25F08" |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:1210) | "TK_PL30W25F16" |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:1428) | "TK_PL30W25F32" |
| (/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib:1647) | "TK_PL30W25F64" |
|-----------------------------------------------------------------------------------------|

Total number of lines parsed: 4101

Resource Usage for Parsing
--------------------------
| CPU Time    | 0.02 sec |
| Peak Memory | 0.10 Mb  |
|------------------------|

Warning:  option "ABSI" ignored


Warning:  option "ABSV" ignored


Source Files
--------------------------------------------------------------------
| "/usr/synopsys/C-2009.09/hspice/bin/bsim3c18rf_to_rev7_2.modlib" |
| "Crystall_Analog_V3_3.lib"                                       |
| "TG03_Gen_nom.sp"                                                |
|------------------------------------------------------------------|

Options
-----------------------
| Option      | Value |
|-------------+-------|
| DELMAX      | 5e-11 |
| METHOD      | TRAP  |
| POST        | 1     |
| PROBE       | on    |
| PURETP      | on    |
| RUNLVL      | 5     |
| SIM_MAXSTEP | 5e-11 |
|---------------------|

Warning: (Crystall_Analog_V3_3.lib:13) resistance limited to 1.000000e-05

46 unique instance models are loaded.

Resource Usage for Load Netlist Data
--------------------------
| CPU Time    | 0.00 sec |
| Peak Memory | 1.29 Mb  |
|------------------------|

License user information
------------------------------------
| FLEXLM   | v8.5b                 |
| USER     | root                  |
| HOSTNAME | localhost.localdomain |
| HOSTID   | 001d608c9d4c          |
| PID      | 30265                 |
|----------------------------------|

License file: @lcserver
----------------------------------------------------------------------------------
| license | checkout number | expiration date     |  # in use / total | server   |
|---------+-----------------+---------------------+-------------------+----------|
| hspice  | 1               | 28-oct-2010/2009.09 | 1/5               | lcserver |
|--------------------------------------------------------------------------------|

License file: @lcserver
-----------------------------------------------------------------------------------
| license  | checkout number | expiration date     |  # in use / total | server   |
|----------+-----------------+---------------------+-------------------+----------|
| hspicerf | 1               | 28-oct-2010/2009.09 | 1/5               | lcserver |
|---------------------------------------------------------------------------------|

Netlist Statistics
---------------------------
| Type            | Count |
|-----------------+-------|
| nmos            | 3     |
| resistor        | 34    |
| capacitor       | 23    |
| vsrc            | 1     |
| diode           | 12    |
| inductor        | 4     |
| mutual inductor | 1     |
|-------------------------|

Summary of Possible Problems in Netlist
-------------------------------------
| Negative resistor             | 0 |
| Negative capacitor            | 0 |
| Negative inductor             | 0 |
| Unconnected node              | 0 |
| Dangling terminal             | 0 |
| Non-rail bulk terminal        | 0 |
| Floating MOS gate             | 0 |
| Floating MOS bulk             | 0 |
| Floating BJT substrate        | 0 |
| Floating controlled input     | 0 |
| Floating current source nodes | 0 |
| Source loop                   | 0 |
|-----------------------------------|

Resource Usage for Netlist Check
--------------------------
| CPU Time    | 0.00 sec |
| Peak Memory | 1.32 Mb  |
|------------------------|

Number of nodes: 42

Total number of nodes (including internal nodes): 57

Resource Usage for Netlist Data Compression
--------------------------
| CPU Time    | 0.00 sec |
| Peak Memory | 1.47 Mb  |
|------------------------|

Simulation

Warning: .ic node N06 does not exist

Resource Usage for Operating Point
--------------------------
| CPU Time    | 0.00 sec |
| Peak Memory | 1.47 Mb  |
|------------------------|

Transient Analysis

9.7% time = 291.0 ns ( etc = 5.0 sec, ett = 5.0 sec )
10.0% time = 300.0 ns ( etc = 5.0 sec, ett = 5.0 sec )
  ( slv=7358  evl=7103  rej=2  mem=1788K cpu=0   wall=0 )
20.0% time = 600.0 ns ( etc = 4.0 sec, ett = 4.0 sec )
  ( slv=13358  evl=13103  rej=2  mem=1924K cpu=0   wall=0 )
30.0% time = 900.0 ns ( etc = 3.0 sec, ett = 4.0 sec )
  ( slv=19358  evl=19103  rej=2  mem=2069K cpu=1   wall=1 )
40.0% time = 1200.0 ns ( etc = 3.0 sec, ett = 4.0 sec )
  ( slv=25358  evl=25103  rej=2  mem=2213K cpu=1   wall=1 )
50.0% time = 1500.0 ns ( etc = 2.0 sec, ett = 4.0 sec )
  ( slv=31358  evl=31103  rej=2  mem=2349K cpu=2   wall=2 )
60.0% time = 1800.0 ns ( etc = 2.0 sec, ett = 4.0 sec )
  ( slv=37358  evl=37103  rej=2  mem=2493K cpu=2   wall=2 )
70.0% time = 2100.0 ns ( etc = 1.0 sec, ett = 4.0 sec )
  ( slv=43358  evl=43103  rej=2  mem=2637K cpu=2   wall=2 )
80.0% time = 2400.0 ns ( etc = 1.0 sec, ett = 4.0 sec )
  ( slv=49358  evl=49103  rej=2  mem=2773K cpu=3   wall=3 )
90.0% time = 2700.0 ns ( etc = 0.0 sec, ett = 4.0 sec )
  ( slv=55358  evl=55103  rej=2  mem=2917K cpu=3   wall=3 )
100.0% time = 3000.0 ns ( etc = 0.0 sec, ett = 4.0 sec )
  ( slv=61358  evl=61103  rej=2  mem=3054K cpu=3   wall=3 )

solve = 61358 eval = 61103 reject = 2

Transient time = 3.87 cpu sec, 3.97 wall sec

Resource Usage for Transient Analysis
--------------------------
| CPU Time    | 3.87 sec |
| Peak Memory | 2.98 Mb  |
|------------------------|

    --- hspicerf done [TG03_Gen_nom.sp] ---

Total time   = 3.89 cpu sec

Total memory = 2.98 Mbytes

\]

ankitgarg0312 · Dec 15, 2009

u have pasted .lis file and .st0 file for 2009.09 versoin.
will it be possible to .lis for both 2004 and 2009 version....

what is freq range ?? , you r using hspicerf here ??

Artemij14 · Dec 15, 2009

frequence range is 1GHz.
I use both hspice2009 and hspicerf2009.
.lis belongs to hspicerf2009 (previous post)
.st0 belongs to hspice2009 (previous post)
They give similar results.

Log files of HSpice2004(was used hspice) simulation in attachment.

ankitgarg0312 · Dec 16, 2009

there are some warning in ur 2009 version lis files...
have a look on them

e.g.
Warning: .ic node N06 does not exist
.ic is very nescsessary to startup oscillations..

Added after 52 seconds:

numbers of nodes are varying in ur 2004,2009 versions ..
are u using same netlist ??

Artemij14 · Dec 16, 2009

Yes, I am using the same netlists.

Despite the warning ".ic node N06 does not exist" the initial conditions are set correctly. Problem is that oscillation is damped. I am afraid that the models of transistors don't work.

Artemij14 · Dec 18, 2009

I solved the problem with the generation of oscillations. The problem was the low Q of resonator.

By increasing L2 of two times I got the generation of oscillations. (Is it correct?I think it should increase the quality factor.)

Maybe my experience will be useful to someone else.

Welcome to EDAboard.com

HSpice 2004 vs HSpice 2009

Artemij14

Newbie level 6

ankitgarg0312

Full Member level 2

Artemij14

Artemij14

Newbie level 6

ankitgarg0312

Full Member level 2

Artemij14

Newbie level 6

ankitgarg0312

Full Member level 2

Artemij14

Newbie level 6

ankitgarg0312

Full Member level 2

Artemij14

Newbie level 6

ankitgarg0312

Full Member level 2

Artemij14

Artemij14

Newbie level 6

Artemij14

Newbie level 6

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Connect with us

Online statistics

Forum statistics